PROJECT SUMMARY/ABSTRACT Patients who survive acute critical illness but require prolonged mechanical ventilation (PMV, i.e. mechanical ventilation >14-21 days) are growing rapidly in number, have large post-acute care expenditures, and experience a grim 60% 1-year mortality. Traditionally, post-acute care of patients receiving PMV has been provided either within the same short-stay hospital or on transfer to long-term acute care hospitals (LTCHs), depending in large part on short-stay hospital practice. It is unclear how a hospital?s tendency to use LTCHs (or not) for post-acute care of patients receiving PMV affects pre-transfer provider behaviors and long-term patient outcomes. Payment reforms, initially intended to enrich for patients receiving PMV at LTCHs (via a tiered Medicare reimbursement model), have led to LTCH closures and potentially also changes in admission and transfer criteria across multiple care settings, including LTCHs, short-stay hospitals, and skilled nursing facilities (SNFs, previously uncommonly used for PMV care). Because certain practice patterns (e.g., early post-acute facility transfer, timing of tracheostomy, feeding tubes, advance directives) occur prior to PMV onset and may determine in part whether a patient undergoes PMV, we will study the immediately ?upstream? cohort in whom such decisions are made: those at high risk for PMV (i.e. mechanically ventilated >96hrs, 1 in 3 of whom will go on to require PMV). We will apply a combination of advanced health services, novel econometric, and qualitative methods to accomplish the following specific aims: (1) Determine trends and drivers of variation in post-acute facility use among patients at high risk for PMV in the 10 years spanning payment reform; (2) Evaluate the association between LTCH use and (a) physician behavior at short-stay hospitals and (b) long-term outcomes of patients at high risk for PMV; and (3) Perform semi-structured qualitative interviews of providers and patients to determine how differences in LTCH use may impact approaches to post-acute care planning and patient experiences. The knowledge generated from this mixed methods approach will be critical to inform both clinical practice (i.e. by altering the way providers plan for and choose optimal sites of post-acute care) and future policy (i.e. by potentially shaping future payment reform and determining whether other sites of post-acute care, such as skilled nursing facilities, should be targets of quality reporting legislation).